Research

The unifying theme of my research is to use patterns in species distributions to understand ecological processes – such as climate tolerances, range shifts, biotic interactions, population dynamics and the timing of life history events (phenology). This means I work at large scales, using models to study hundreds of species across countries and continents. I actively use this research to inform conservation policy in our changing world, evaluating the effectiveness of current and proposed conservation strategies.

The ecological niche

Thousands of species, such as this mustard, have been introduced into the USA. I am interested in how they are able to survive in climate conditions that they do not occupy in their native distribution.

Why do some species only live in a few places, whilst other species are found across entire continents? Species can tolerate different environmental conditions, their distributions may be limited by competition or predation, or by geographic barriers such as mountain ranges. However, there is little consensus on the relative importance of each of these factors. Resolving this is crucial now more than ever, because we need to predict how global change will affect species distributions.

Agrimonia eupatoria was introduced into North America as a herbal remedy. Historic data on species introductions from archive sources, has helped evaluate how often models of species distributions are wrong

If the species distribution models we use to measure the environmental conditions species can tolerate (the fundamental ecological niche) are wrong, then the predictions we use to plan conservation will be wrong. We need independent data on species distributions to test our models. I am using the distributions of many hundreds of species in areas where they have been introduced outside of their native range in order to do this. During my post-doc with Dov Sax I collected a database on the native and exotic distributions of >2000 plant species – the most comprehensive and complete database of its kind that we know of. Remarkably, our results indicate that models of environmental conditions alone misestimate the locations species can live in by a substantial degree.

So non-environmental factors are very important for determining species distributions. The next question is which of these factors can we measure and include in distribution models? To address this I am collaborating with a range of scientists internationally, including Dr. Oscar Godoy (UCSB, USA),  Dr. Miguel Araújo (Universidade de Évora, Portugal), Dr. Fernando Valladares (Natural History Museum, Madrid) and Dr. Dov Sax (Brown University, USA).

Climate paths

Taricha torosa, the California Newt, is likely to be in trouble because under climate change its Californian coastal range is becoming climatically unsuitable for it. In order to survive the newt may need to move hundreds of miles inland to the Sierra Nevada, crossing the agricultural region of the Central Valley.

Climate change is making the habitat of many species unsuitable, forcing wildlife to move to new places in search of suitable habitat. Many species won’t make it without our help. To do this we need to know the routes along which species are going to move, i.e. ‘climate paths’. I have been studying how population processes such as colonisation and extinction will affect how amphibians in California move along their climate path (press coverage here). This research found that even in the absence of geographic barriers, climatic fluctuation can lead to gaps in the climate path that will prevent range shifts. Short term population persistence under unfavourable conditions can be critically important to achieving range shifts. I am now studying the effects of agricultural and urban habitat fragmentation on the climate paths of vertebrates in the Iberian Peninsula. The goal is to figure out the best way to protect species affected by climate change – habitat corridors or managed relocation?

Population dynamics and conservation planning

Populations are dynamic and they move around. One year a species may  be thriving in a nature reserve and the next year they may be doing much better somewhere else. So we can’t just choose one place to conserve species. What many species need is networks of reserves, so that if one population disappears, it can be recolonised from somewhere else.

The Marsh Fritillary is one of the most endangered butterflies in Europe, but it has a stronghold in Wales.

The first thing we need to plan reserve networks is a complete map of all the places that populations of a species might live. Using distribution models at really high resolutions I made a habitat map for the Marsh Fritillary butterfly (Euphydryas aurinia) in Wales, UK. This map did a great job of identifying the important conservation areas for this endangered butterfly (published here). Having a map like this enables us to do something unusual – simulate population dynamics across a large region. Most habitat maps are usually only made for small parts of a species distribution, encompassing a few populations. Simulating population dynamics for hundreds of populations gives us a new insight into how these populations work together. We expect that having large populations in areas where habitat is abundant is important for a species to survive. But it appears that small, isolated populations can link together these large populations, helping recolonise them if a local extinction occurs. These small populations contribute disproportionately to species persistence and should be conserved as well as large populations.

Zonation output for butterflies in Britain. The red areas are the most important landscapes for butterfly conservation.

Although to figure out where to put reserve networks we ideally want detailed models of population demography, spending years getting these data is unrealistic given the immediate threats species face. So we need to use proxies. I have worked on using ‘connectivity’ between populations as a proxy for survival probability. I collaborated on the development of a new algorithm ‘Zonation’ that uses connectivity maps for dozens of species simultaneously to pick out the best sites for reserves for all native butterflies in Britain (published here).

But if we are to trust techniques like Zonation, we need to know if its reserve design will really ensure that species persist. I tested whether a multi-species solution can really do a good job of protecting individual species (published here). Zonation turns out to be quite effective at picking the landscapes where rare species are most likely to persist in the long term. But we learn the lesson that the political choices regarding species ‘importance’ can have strong effects on how well species are protected.